Does the presence of professional football cause crime in a city? Evidence from Pontiac, Michigan

ABSTRACT

Using daily panel data from Detroit, we empirically explore the relationship between the National Football League (NFL) and crime in a city. We exploit the natural experiment of the Detroit Lions’ move from Pontiac, Michigan, to downtown Detroit in 2002. Pontiac is used as the treatment city and non-game day crime, other suburban cities, and other cities outside Detroit MSA are used as the comparison groups. Employing a difference-in-difference approach, we find decreases in assaults and vandalism on home game days in Pontiac relative to the control cities after the Lions’ move. We find weak evidence of a net decrease in larceny and vandalism in Pontiac on home game days following the loss of professional football. No changes in assaults and auto theft are reported. While not conclusive, our results suggest that professional football leads to additional larceny and vandalism incidents but no effects on assaults and auto theft.

KEYWORDS:

• National Football League
• stadium
• National Incident-Based Reporting System
• crime
• externality

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Two other outcomes that would be valuable to look at include car crashes or DWI citations. These are classified under Group B offences in the NIBRS data. We follow Kalist and Lee (2016) and only look at Group A offences since law enforcement agencies are required to report both incidents and arrests for Group A offences, but only arrests for Group B.

² Yu et al. (2016) find that home college basketball games are associated with robberies.

³ The finding that spending on professional sports is offset by reducing spending elsewhere in local economy can be found in a large number of studies. See, for example, Rosentraub et al. (1994), Coates and Humphreys (2003), and Coates and Humphreys (2008).

⁴ The field of sports economics is a relatively young and empirical (Hall, Humphreys, and Pyun 2017). See, for example, Caudill (2009), Swofford, Mixon, and Green (2009), and Cebula, Toma, and Carmichael (2009) and other papers in the special issue of the applied economics of sport edited by Taylor (2009).

⁵ Some cities in Detroit MSA, such as Detroit and Livonia, are not included in this comparison group as they do not report crime data to the NIBRS during the sample period.

⁶ Note that daily, monthly, yearly, and city-specific fixed effects are jointly significant but not reported in the interests of space.

⁷ Similar to the results of Table 3, all coefficients for daily, monthly, yearly, and city-specific fixed effects are jointly significant, respectively, though they are not shown in Table 4.

⁸ We say weak, because we find only one supporting evidence from some analyses.